Self antigens, in the form of differentiation antigens, and altered self antigens, in the form of mutations, are recognized on melanoma and other cancers by the immune system. Although mutations are an accepted basis for tumor-specific antigens, we and others have shown that recognition of differentiation antigens is also relevant to immunity to cancer. Recognition of these antigens presents problems, especially how one can overcome immune tolerance. We have used a mouse melanoma model to show that mice can be tolerant to the tyrosinase family of differentiation antigens. Tolerance can temporarily be broken by immunization with xenogeneic orthologues of antigen. However, our preliminary results suggest that a CD4+ cell population and B cells suppress immune responses to melanoma. Our first aim addresses the role of CD4+ and B cells n suppressing tumor immunity and memory responses, particularly to different action: antigens. Our next two aims turn to the consequences of immune recognition of mutations. We have developed methods to create combinatorial libraries of mutated differentiation antigens to address what types of mutations can lead to autoimmunity against the non-mutated parental antigen (aim 2) and what type of mutations can lead directly to tumor immunity (aim 3). In aim 4, we will continue to create more relevant mouse models of melanoma that develop endogenous invasive melanomas to study tumor immunity against differentiation and mutant antigens.